Abstract:

There are universal expectations of basic skills a student will gain in computer science education. For instance, students need to understand how the logic of programming works, including knowing more than one paradigm and programming language. Despite these universal expectations and the best intentions of guidelines, such as ACM/IEEE Curriculum Guidelines, there is no one size fits all approach to teaching introductory computer science courses, and the result is a diverse range of teaching and evaluation methods.

The goal of this paper is to look at the varying approaches universities have taken to teach introductory Computer Science (CS) courses and evaluate student’s knowledge. By conducting a systematic literature review on what institutions have done when faced with the issue of designing introductory courses, we compiled relevant information that is of substantial value. This framework includes an analysis of university type, class size, intended student body, programming language, assessment methodology, course contents, and the quality of the paper. This paper aims to improve the understanding of how introductory computer science courses are being taught in order to better guide the construction of introductory courses in the future.

This systematic literature review began with an analysis of 1,741 peer reviewed papers published in the past 15 years and 56 of these papers were accepted for this research. Our findings confirmed our assumption that there is not a singular best practice method in which introductory computer science courses are conducted. We did find certain methods have been more frequently used with more success than others in the past 15 years. Most commonly introductory courses have used Java or Python as programming languages. Additionally, during this time span there has consistently been offerings of different introductory courses based on a students academic background such as being a CS major, a non- CS major, or a different kind of STEM major. There are some trends that have gained relevance over time and now are among the most common methods of intervention in CS classes: pair programming and project based intervention. Our findings show the successful intervention models in use by universities around the world, and in general in which context they were used, as a way to guide instructors what interventions have a positive record in their specific context.

Keywords:

Introductory Programming, Curriculum, Systematic Review.